One of the more puzzling aspects of our taxpayer-funded Met Office is its apparent ability to confidently predict local climate several decades into the future when it is seemingly unable to predict our weather from one day to the next (you’ll be surprised to learn that their climate forecast predicts that it’s going to be hot – although this prediction was made before the Mann/Trenberth/Gore paper, peer-reviewed by Louise Gray and Johann Hari, which conclusively demonstrates that ‘hot’ actually means ‘cold’.. unless it’s a hot day, of course).
Autonomous Mind (AM) has an admirable record with regards to exposing the shortcomings of the Met Office and, in particular, its complicity in the climate scam: for example, see this group of writings about their 2010-2011 winter ‘forecast’.
Inevitably then, AM has attracted many comments about weather forecasting and last Sunday, he posted this;
In previous comment threads over recent months there have been several references to weather forecasts adjusting significantly over the five day lead time. So this blog is carrying out its own, albeit unscientific, ‘weather test’ to compare forecasters’ predictions over the next five days for the weather on Thursday.
A number of others volunteered to do something similar: these are my observations for the Eastbourne area.
I used the same forecast sites as listed by AM but because I’ve noticed forecast variability over the course of a few hours – as the forecasters fit their predictions to real-life observations – I chose to make screengrabs twice a day, morning and evening (except for the Sunday which was evening only). For reasons of space – and to avoid too much repetition – I have only included screengrabs that I feel are of any significance: should anybody want to see them all, send me an e-mail and I will happily make them available.
As I write this post at about 4pm, the Eastbourne weather has been a mix of sun, cloud and quite stiff winds. Temperature has peaked at 15ºC and we’ve had a couple of light showers – one at approximately 7am and another at about 3pm.
On the first day of monitoring, Sunday 3 July, all the forecasts predicted a mix of sun and showers with a maximum temperature in the region of 15ºC. Weather Outlook over-estimated the extent of morning rain – but their site does break the weather into three-hourly periods while most of the others merely predict showers at some unspecified time.
All in all, each of the sites on Sunday provided a reasonable forecast for today’s weather.
Thereafter, Weather Underground and Accuweather forecasts remained fairly consistent throughout the week. Because they provide more detail, the other sites had more opportunities for discrepancies – the main problem being the extent of rain that they were forecasting. If I had been reliant on Weather Outlook, the chances are that I would not have planned to be outdoors today – whereas the day has proved to be almost completely dry.
In fairness, none of the forecasts have strayed too far from their Sunday position although the BBC on Tuesday morning was briefly predicting thunder for today.
I’d venture the opinion that of all the forecasts, the Met Office was the least accurate but this is only on the basis that it stopped predicting rain altogether on Monday and Tuesday; by Wednesday evening, it suggested heavy rain at about 1pm. This morning, it was not showing the rain that was falling at the time and throughout the day, it has been pushing the arrival of rain back by hours at a time.
As AM said, this is not a rigourous scientific exercise and, overall, I’m genuinely surprised by the degree of accuracy contained in the forecasts. However, previous experience suggests that this may be an exception and like all good climate scientists, I will repeat and tweak until I get the result that I first thought of!
Update – As I’m about to hit the ‘publish’ button at 5.45, it is has started to rain heavily.